Is it possible to precision rectify a 1MHz signal?

Question

For an Overcurrent detector, the present solution is a full wave bridge of MBR0530 diodes, and the forward drop of these diodes is the cause of increasing error towards the low setting. 
 The current signal is presented across a 5.1 ohm resistor from a 1:1000 current transformer, rectified, and compared to an adjustable and metered reference, with an AD311 comparator. 
 There is a 5 volt and a 9 volt supply available, and the frequency of the signal can vary from 20kHz and up to 1 MHz. 
 Is it possible to build a precision rectifier that runs from a single supply, and functions up to 1 MHz? 
 Cheers, Finn Hammer

FinnHammer · Accepted Answer

It is always nice to be able to answer one's own questions, since it usually means, that the effort to find out has been succesful. I tried to build the circuit mentioned in this publication: 
 http://www.analog.com/library/analogDialogue/archives/44-04/absolute.pdf 
 but instead of the AD8277 I used AD8028 and 2.2k resistors instead of 40k, I got this result at 500kHz, which is sort of ok, I think 
 At higher frequencies, the inverted part of the signal gets more ragged, and obtains a peak at the leading edge, This peak exeeds the absolute value around 800kHz, and this is exactly within speck, so I will stick with this solution. 
 
 Cheers, Finn Hammer)

Electronic Design Services - PartnerZone

Products Mentioned

Is it possible to precision rectify a 1MHz signal?